Edited Property Briones

Property Rights Reform in Philippine Agriculture: Framework for Analysis and Review of Recent Experiences Roehlano M. Briones

In traditional and modern economies, an essential element of the resource allocation mechanism is the property rights system. Flaws in the underlying property rights system may lead to misallocation of resources. The related problems of mass poverty, sluggish growth and rapid resource depletion in many developing countries suggest a need for allencompassing property rights reforms. The Philippines exemplifies these problems: the official poverty headcount in 1997 was 37.5 percent, with a far larger magnitude (51.2%) in rural areas, where land ownership inequality is common. Deforestation and overfishing remain critical: in 1991 to 1996, the average annual rate was 1.52 percent. The forest cover was 5.54 million hectares, down from 10.8 million hectares just three decades before 1968. Meanwhile, municipal fisheries have experienced an average annual output decline of 2.8 percent in the 1990s. The agriculture and natural resource sectors have therefore been the subject of extensive property rights reforms. The Comprehensive Agrarian Reform Program (CARP) began in 1988 to implement RA 6657 (the Comprehensive Agrarian Reform Law, or CARL), a culmination of several decades of land reform efforts. These efforts include the land reform code (RA 3844) of the 1960s, which prohibited share tenancy, and the 1972 Presidential Decree No. 27, which covers rice and corn lands. Meanwhile, policies on property rights in the uplands and coastal areas have been implemented to arrest resource degradation, including the introduction of tenure instruments and the recognition of community-based management. In the past decade, studies on these initiatives have burgeoned. This paper reviews these studies using an analytical framework focusing on efficiency and equity. Interpreted broadly to include intergenerational welfare, these criteria subsume the issue of

2 sustainability. Economic analysis typically focuses on efficiency, for which its analytical tools are well suited. The equity criterion, while important, involves a nearly intractable web of social justice issues. This review finds that focusing on the efficiency criterion is fruitful for evaluating property rights reform. Equity nevertheless remains integral to the discussion; as one can see later in this discussion, recent theories hypothesize complementarity (rather than tradeoff) between efficiency and equity. Evaluating efficiency involves a comparison of benefits and costs. The challenge to an efficiency analysis is in identifying and valuing such benefits and costs of reform. Given the volume of relevant existing literature, this paper devotes the most attention to the CARP, which is the Philippines’ premier property rights reform program.

Analytical Framework For Land Rights

Property rights systems

Property rights determine how persons may or may not behave with one another, with respect to the use of resources. A system that assigns property rights confers obvious benefits to society as a whole. First, in the absence of such a system, a destructive “state of nature” may prevail, wherein individuals and groups engage in conflict to appropriate one another's possessions. Second, the property rights system enables a reward-andpenalty scheme to operate, motivating persons to employ resources toward productive ends. On the other hand, property rights systems also impose costs. At the very least, there are transaction costs in setting up and maintaining the system. For example, resources are used when demarcating and defending boundaries, and establishing a mechanism to settle conflicts and adjudicate disputes.

3 Private property

The most familiar property rights system is that based on private property. In this system, property is defined in terms of residual rights1, which permit the rights holders to take actions not otherwise prohibited within the system. Discretionary use, within explicit restrictions, is the most natural way of understanding the term “right”. An obvious example is the ownership of livestock with prohibitions on cruel treatment. In addition to assigning residual rights, the system allows the transfer of these rights. Transferability is a critical feature of the private property system, as this permits the operation of markets, (which are simply institutions for the orderly exchange of rights). The formation of markets is explained by the evolutionary theory of property rights (Fields 1989; Platteau 1996). The classic statement of the theory is attributed to Demsetz (1967): “Property rights develop to internalize externalities when the gains of internalization become larger than the costs of internalization. Increased internalization results from changes in economic values" (p.350). He cites as an example the development of land rights for Labrador Peninsula Indians: with increased fur trading came the formation of a property system in forestland. The evolutionary theory also appears in Boserup (1965). The author identifies population pressure with technological change leading to settled agriculture as factors that help define private property rights more accurately. A more recent illustration of the theory is the case of Thailand, where the opening of the Kingdom to international trade in the 1800s increased demand for rice land, thus ushering in a formal land registry system (Feeny 1982, as cited in Deininger and Feder 1998.) The implementation of land rights has a growing effect on investment (Deininger and Feder 1998). In China, the shift from collective farming to individual land rights raised

1

The term is appropriated from incomplete contracts literature; see Grossman and Hart (1986), and Hart

and Moore (1990).

4 agricultural productivity (Lin 1992). Case studies of a communal forestland show that the establishment of individual land rights motivated agro forestry investment (Otsuka 1998). Active participation of farmers in Cameroon in land titling programs is additional confirmation of the incentive effects of land rights (Firmin-Sellers and Sellers, 1999).

Agricultural land under a private property system: Allocation and welfare

The following now clarifies the criteria to be used in assessing social welfare and then identifies the welfare implications of the operation of a land market. Evaluation criteria

A widely used (and fairly innocuous) standard for evaluating alternative allocations is the Pareto criterion. The criterion states that a reallocation makes society better off if it harms no one, but improves someone’s well being. Social welfare is maximized if no further Pareto improvements are possible, in which case the allocation is characterized as Pareto optimal. This implies that the allocations that maximize welfare may not be unique; rather, that there may be a range of optima corresponding to a range of possible initial allocations of individual well-being. A more useful but far more controversial standard is the criterion of potential Pareto improvement. According to this standard, reallocations that harm another person makes society better off if there is a compensation scheme financed by beneficiaries such that nobody ends up being worse off. As with the Pareto criterion, optimality is defined as exhausting all such improvements, and a range of welfare optima exists consistent with alternative initial distributions. The controversy surrounding this criterion is due to the purely hypothetical nature of the compensation. Payment is seldom made given the costs of actually setting up a

5 compensation scheme. Despite this difficulty, the standard is widely used in normative analysis and is employed in this paper.

Land use under perfect markets

This section now clarifies in a more precise manner the benefits that arise from the private property and exchange system. The benchmark case of perfect markets is first examined. As defined in neoclassical theory, this case refers to an idealized state where transaction costs and externalities are absent, agents behave competitively, entry and exit choices are unconstrained, and information is equally distributed across economic agents. Two major propositions may be stated for this case:2 First, market equilibrium is optimal. Optimality of land use may be understood in both a static and dynamic sense. From a static viewpoint, payment for land services would be the same for all transactions at equilibrium and would equal the marginal product value in all uses. Hence, no reallocation of land can increase aggregate output. Other factors are priced and employed in the same way. From a dynamic viewpoint, landowners will make optimal choices on land improvement (or preservation of land value) as they undertake land investments until the point where the rate of return equals the opportunity cost of capital. Moreover, the price of the land capitalizes the present value of its income stream. Second, any optimal allocation corresponds to some market equilibrium. That is, the attainment of a Pareto optimum does not depend on the distribution of property rights. The implication is that a preferred outcome, based on a more restrictive welfare criterion (e.g., one that emphasizes equity), may be reproduced as a market equilibrium starting from some appropriate endowment. To reach this appropriate endowment, some authority

2

These are of course the two fundamental welfare theorems.

6 may have to undertake redistribution aimed at endowment values or individual assets as a whole; land commands no special attention as a store of wealth. The caveat is that redistribution must be costless. Of course, this is rarely true; those harmed by redistribution may expend resources to oppose the move, as well as make choices that would be otherwise be unprofitable in the absence of intervention. That is, the intervention can introduce distortions. For example, a tax-transfer system may reduce labor supply of both contributors and recipients. Because of these costs, a move to a greater equity may face the efficiency-equity tradeoff.

Land use in the real world

In reality, markets, particularly in agriculture, are far from perfect. Transaction costs abound; contract enforcement may be ineffective; information is asymmetric; competition is imperfect; and some markets may be absent. Agriculture is moreover subject to seasonal fluctuations and random shocks that exacerbate these problems. In the presence of such market failures in agriculture, the foregoing propositions are no longer valid. Market equilibrium need not be efficient, and the attainment of a Pareto optimum may actually depend on the initial endowment. The inapplicability of the second proposition above requires a new approach toward the analysis of efficiency and equity. Instead of a potential efficiency-equity tradeoff, the simultaneous improvement of efficiency and equity may be possible. Theoretical explanations for the equity-efficiency link in land distribution may be the following: (1) imperfections in credit markets (2) agency relations in land cultivation (3) localized linkages for income diversification .

7 The efficiency-equity link: Imperfections in related markets

Failures in the operation of land markets may arise from imperfections in related markets, particularly the credit market. Credit constraints are the most commonly-cited and convincing argument for the inefficiency of skewed land ownership (Deininger and Squire 1998; Aghion, Caroli and Peñalosa 1999). The credit problem is the consequence of asymmetric information and weak contract enforcement. An investment project always involves risks exogenous to the credit market (e.g., pest infestation or steep price declines). Inadequate project returns are also related to the behavior of the borrower, who may decide to default, take excessive risks or fail to exert the effort needed to ensure a profitable return. Such behavior may be the result of the inherent individual characteristic: reliability. The "reliable" and "unreliable" types have a known distribution in the population, although information on the type of a particular borrower is not known to the lender. When the interest rate is a signal of the type of borrower, credit may be rationed in equilibrium (Stiglitz and Weiss 1981). The problem of price being related to quality (i.e., interest rate on a loan with the assurance of repayment) is referred to as adverse selection. On the other hand, default or the potential to default may be the borrower's choice and therefore not an external cause. Information regarding the choice taken is asymmetric. This problem is referred to as moral hazard. Market adjustments to these problems are evident in the agricultural sector. Credit constraints artificially raise factor supply for activities that generate immediate cash flows; for example, farm households allocate labor supply away from own-farm labor to off-farm labor, leaving part of their cultivated area idle (Alwang and Siegel 1999). In the formal sector, the collateral system is a common response, with land as the favored form of collateral given its immobility and relative ease in ascertaining property rights and its transfer. However, actual land prices will probably be pushed above the capitalized net yield of the land because of its collateral service (Binswanger, Deininger and Feder 1995). In the informal sector, credit may be tied in with other transactions, such as the purchase of harvest (Floro and Yotoupolos 1991).

8 With imperfect credit markets, households can easily fall into “poverty traps”. This is exacerbated by the presence of random shocks on assets (e.g., crop loss, medical emergencies) to which the household is vulnerable under credit rationing and in the absence of insurance markets. Once the household falls into the poverty trap, they are unable to obtain financing for the investment whose returns could pull them out of the trap. A simple model proposed by Banerjee and Newman (1994) illustrates how constraints to investment credit can generate such traps. The model generates credit rationing given the moral hazard problem. Suppose only one good is produced, according to a production function f(k), where k is the amount of capital; as usual, f ' > 0, f " < 0 . To purchase k, the

individual borrows the amount, pays the interest rate equal to r, and puts up his current wealth w as collateral. The borrower may, however, abscond, forfeiting w. If he evades creditors, the person keeps f(k); if he is caught, he gets nothing. The probability of being caught is π (k ), 0 ≤ π (k ) ≤ 1, where π is nondecreasing in k. The agent is risk neutral; hence, he seeks to maximize expected profit. Incentive compatibility requires

f (k ) − (k − w)r ≥ [1 − π (k )] f (k ) . Optimal borrowing, denoted as k * , is at f ' (k * ) = r. Then solve for critical wealth w* that just provides a repayment incentive at the optimal investment:

w* = k * −

π (k * ) r

.

If w falls below w* , then the lender does not lend the optimal investment level. The reason is that there is a lower bound (zero) on the borrower’s punishment. As the poor are

9 already near zero wealth, then they are less deterred by the prospect of being caught. Note that if an individual has an endowment greater than w* , and another slightly below it, then a reallocation of capital or wealth from the former to the latter in the current period will increase aggregate output. Information and enforcement problems prevent the market from making this reallocation. The efficiency-equity link: Agency relations in land cultivation

Another source of failure in land markets arises through agency relations in land cultivation. When ownership of land is skewed, the owner inevitably delegates production tasks to agents. Two options are available: first, the owner hires and supervises wage labor; second, the owner delegates operations to a tenant-cultivator. For the first option, the difficulty lies in the high cost of supervising wage labor (Family labor requires less supervision but of course is in limited supply). The transaction cost of using wage labor is the generally explanation behind the pattern of declining yield that accompany farm size increases (Faruquee and Caree 1997). Doubts have been raised against the robustness of this stylized fact. For example, Benjamin (1995) suggests that the pattern may be partly due to the omission of variables such as soil quality. The relationship is reasserted by Heltberg (1998). Hayami, Quisumbing and Adriano (1990) confirm the absence of clear empirical evidence regarding economies of scale in agriculture. In the case of plantation agriculture, the apparent scale economies can be attributed to coordination problems at the processing and distribution stage, and not from increasing returns at the production stage. The second option, which is to delegate operations to a tenant-cultivator, takes the form of either fixed rent or sharecropping. Sharecropping, however, faces an incentive problem if the agent receives earnings that are lower than the marginal product value of the land. In contrast, the cultivator under leasehold receives this marginal product value. The "Marshallian inefficiency" hypothesis posits that the effort of the agent is lower under

10 sharecropping than under leasehold. Share tenancy is regarded as a feudal vestige to be superseded by leasehold as the countryside begins to progress economically. However, the prevalence and persistence of share tenancy has brought forth theories that examine its basis in rational contracting. Tenancy contracts must combine work incentives with risk-sharing. If the agent shirks from working due to high transaction costs involved in the monitoring process, any output losses from such shirking (a form of moral hazard) cannot be distinguished from those due to environmental factors. Sharecropping, on the other hand, may be used as incentive to workers who also share the risks in agriculture (Stiglitz 1974; Newberry and Stiglitz 1979; Otsuka and Hayami 1988). An interesting suggestion by Eswaran and Kotwal (1985) distinguishes two types of moral hazard: in addition to shirking by the cultivator, the landowner may also shirk in providing managerial services to production. This provides an additional justification for the share contract. In their model, depending on the specific set of local conditions and agent preferences, sharecropping may be Pareto superior to leasehold. Instead of moral hazard, another form of asymmetric information that could lead to sharecropping is adverse selection (Hallagan 1978; Muthoo 1998). Own skill or ability may be the private knowledge of a farmer; hence, the contract offers are designed to screen farmers based on

skill. However, the adverse selection explanation for

sharecropping does not seem to jibe with evidence that finds information regarding skill to be public knowledge within a village (Lanjouw 1999). Moral hazard is deemed to be the more likely explanation for share tenancy. The concept of sharecropping also came about as an outcome of an imperfect labor market (Ray 1999). In this interpretation, share tenancy is a form of strategic delegation undertaken by competing landowners. The model is set up as follows: (1) The landlord can either operate the farm and hire workers, or delegate operation to a share tenant.

11 (2) The structure of competition is such that when one farm reduces its wage, a rival farm must also reduce its wage. (3) The decision to hire a tenant is construed as a credible commitment that a lower wage will be paid, because the tenant is given only a partial output incentive. This view is hospitable to the thesis of Marshallian inefficiency.

A survey of the empirical literature by Hayami and Otsuka (1993) nevertheless yields little evidence for Marshallian inefficiency. Agrarian contracts are found to adapt to real world enforcement problems; hence, share tenancy is more frequently observed in cases where monitoring is less costly, i.e. in closely-knit communities and families. A study for the Philippines confirms that share tenancy contracts between kin did not weaken production incentives (Sadoulet, de Janvry and Fukui 1997). There seems to be no compelling reason to attribute production inefficiency to sharecropping. There are on the contrary, good reasons to believe that sharecropping promotes efficiency, based on the foregoing. It permits cultivation by family labor, reduces reliance on hired labor, spreads risk between tenant and landlord and provides incentives for landlords to supply managerial input. Sharecropping also mitigates credit problems, as sharing of output is often accompanied by sharing of cost outlays. Often, landowners purchase fertilizers, pesticides and other inputs, and share their cost (Otsuka 1999). Moreover, the harvest rights acquired by the tenant upon gaining access to land opens up credit opportunities from moneylenders and traders. These advantages imply that share tenancy promotes equity. Given the traditional criticism of sharecropping as a brutal form of exploitation, the revisionist view of modern economic theory is quite remarkable. The efficiency-equity link: Localized linkages in income diversification

The final argument regarding the inefficiency of inequality rests on the potential for localized linkages to promote rural industries when incomes are evenly distributed. "Localized" is used here to distinguish this from the traditional view that sees land reform

12 as a means to indirectly promote national industrialization. According to this argument, industrialization is precluded by the excessive diversion of productive capital into relatively unproductive assets of the landed class (see e.g., Cornista et al. 1989). There is no coherent framework to account for this persistently inefficient and perhaps irrational behavior, nor is there persuasive evidence to show that urban-based industrialization requires liquidation of the landed class' assets. A localized linkages framework on the other hand uses the following argument: Consider a rural economy divided into a household and a production sector. The latter is further subdivided into agriculture and manufacturing. The development of rural manufacturing is initially dependent on strong local demand, both from the agricultural and household sectors. The distribution of household incomes may be a determinant of this demand when nonfarm goods are income elastic. The argument is most applicable to an agriculture-dependent region where the greater bulk of the population is poor, and land ownership is highly skewed. Agricultural development and asset redistribution may be the impetus for rural growth linkages (Ranis and Stewart 1993). The importance of income equality in promoting local linkages is echoed by Park and Johnston (1995) in the case of Taiwan. The concept of linkages can make sense3 by relating it to the concept of scale economies (Krugman 1993), assuming a case of perfect markets. Fafchamps and Helms (1996) construct a formal model of local linkages, summarized as follows: consider a village whose transactions with the outside world are costly. Suppose rural manufacturing is characterized by increasing returns; moreover, the proportion of manufactures in total expenditure rises with income (i.e., preferences are nonhomothetic). A multiple equilibrium situation is possible, with some levels of Pareto equilibrium dominated by

3

In a rural setting, linkages are conceptualized in a Keynesian type framework, and are quantified by

computing localized multipliers. This approach, however, lacks micro foundations (as does the macro model from which it derives), and its appropriateness for explaining development per se is suspect. See Briones (2000).

13 others. In a low-level equilibrium, village manufacturing is underdeveloped because of low demand for its output---but demand is low precisely because incomes are low in the absence of highly productive manufacturing activities. Under these circumstances, how a costless asset redistribution affects the size of the rural manufacturing sector depends on the composition of output demand as income varies. In one of their simulations for Guatemalan villages, the researchers of the study found that redistribution efforts meant to reduce asset inequality could trigger rural small-scale industrialization. These arguments hint that the industrialization and development process is dynamic. Likewise, the efficiency-equity link model can be completed only by looking at its dynamic aspect.

Dynamic version of the efficiency-equity link

The dynamic version of the model of Banerjee and Newman (1994) concludes that a onetime wealth redistribution can alter the equilibrium path and lead to higher rates of longrun growth. That is, redistribution increases the wealth of the poor and permits them access to credit markets, which increases the over-all rate of capital accumulation. The efficiency-equity link may be stated in dynamic terms as the growth-equity hypothesis: a more equal asset distribution leads to higher future growth rates. In contrast, the traditional view is expressed by the Kuznets hypothesis: as an economy grows, inequality at first rises before declining. Recent evidence tends to support the equity-growth hypothesis (Alessina and Rodrik 1994; Clarke 1995). In particular, Deininger and Squire (1998) zero in on land inequality as a determinant of future growth. The distribution of operational landholdings is a proxy for asset distribution. Using a pooled time-series and cross-section country data (characterized as a “high quality” data set), they find little evidence for the Kuznets hypothesis; instead, there is a strong negative relationship between inequality and growth. Moreover, weak

14 income growth tends to be concentrated at the lower end of a skewed income distribution.4 The theory and available evidence seem to favor a one-time redistribution of assets for an economy with marked inequities and feeble growth. Where landholding is an important store of wealth and favored collateral form, redistribution may be targeted at landowning. However, this one-time redistribution of land may not lead to a permanent reduction in land inequality once risk is introduced into the analysis. Random shocks on individual wealth holdings, when credit and insurance markets are imperfect, may result in widening gaps between the lucky and the unlucky over time. Co-variation of shocks, localization of land markets, and the absence of nonfarm employment reinforce these distributional trends. Land sales tend to be concentrated in periods of adverse natural or economic conditions (e.g., drought or low prices), where “distress sales” within an area (e.g., stricken by a common shock) force land prices down. Land buyers, on the other hand, possess financial assets. Hence, there may be a secular tendency for land ownership to continue to be concentrated (Binswanger and Deininger 1997; Carter and Zimmerman 1998). This consideration may justify complementing land redistribution with programs that offset adverse shocks such as credit support. The following section therefore now examines the case of the CARP in the Philippines, which combines land redistribution with support service provision.

4

The equity-growth view is not without its detractors. Li and Zhou (1998) argue that, if

public

consumption goods are financed by a growth-reducing tax, then majority voting leads to a distortionary regime when inequality is low. A regression of GDP growth on past values of the Gini co-efficient confirms the hypothesis.

15 The Comprehensive Agrarian Reform Program

The CARP is arguably the most ambitious program for property rights reform in the Philippines. In principle, the CARP encompasses all agricultural lands. The actual coverage is estimated at 8.06 million hectares, or around 83 percent of agricultural lands. Of these, 4.32 million hectares (around 54 %) are private lands, government lands and resettlement areas, all falling under the jurisdiction of the Department of Agrarian Reform (DAR). The remainder (3.74 million ha.) consists of public agricultural lands, including public alienable and disposable lands and some forest lands, falling under the Department of Environment and Natural Resources (DENR).

The CARL sets a 5-hectare ceiling on private land ownership. An additional 3 hectares for each of the landowner's children may be retained for owner-cultivation. Land sizes exceeding these retention limits are acquired by the Land Bank of the Philippines (LBP). Landowners must be paid a "just compensation" to be determined in the course of implementation. Up to 35 percent of the compensation may be given in cash, with the remainder in the form of government financial instruments and LBP bonds. The yield is set by the 91-day T-Bill. Landowners may also opt for “voluntary” sale or transfer. Ownership of acquired land is then transferred to cultivators. Persons prioritized to receive transfers are tenants, followed by regular farmworkers, seasonal farmworkers, other farmworkers, tillers of public land and other cultivators. Agrarian Reform Beneficiaries (ARBs) are entitled to no more than 3 hectares of land. In the case of acquired private land, ARBs are required to pay 30 annual amortizations to the LBP at 6 percent interest. Exempted from coverage are lands for public use, livestock and poultry farms, prawn farms and fishponds, and lands converted to nonagricultural use. The last exemption has gained notoriety, given the Local Government Code provision that authorizes local

16 governments to reclassify up to 15 percent of agricultural land in their jurisdiction. In addition, compulsory acquisition of commercial farms will be deferred for 10 years after the validity of the program. For commercial estates, CARP provides alternative arrangements for asset reform such as stock distribution and profit-sharing. Land acquisition and distribution (LAD) is scheduled by the CARL as follows: Phase 1 (1988-1992): Rice and corn lands under PD 27, idle lands, private lands under

voluntary sale or transfer and government lands; Phase 2 (1988-1992): Public alienable and disposable lands, resettlement areas and

private agricultural lands in excess of 50 hectares; Phase 3 (1992-1995): Private lands between 24 and 50 hectares; (1994-1998) private

lands below 24 hectares. Other features of the program include prohibitions against sharecropping as well as rental ceilings for leasehold, which are carryover policies of previous land reform legislations. Moreover, land transactions are effectively frozen: land covered by CARP cannot be sold while distributed land cannot be transferred for the next 10 years. (The exceptions are transfers to the government or the LBP, or transfers by inheritance). Finally, the CARP seeks not only to redistribute land but raise agricultural productivity by providing support services to ARBs as well. These services include the provision of credit, infrastructure, technical assistance and community organization. To operationalize beneficiaries’ development, the CARL also provides for the creation of Agrarian Reform Communities (ARCs). These are each composed of a barangay or a cluster of barangays and “primarily composed and managed” by ARBs. In each area, a farmer’s organization or cooperative will be identified to take the lead in the agricultural development of the locality. The ARC embodies in principle the development approach anchored on participation, local empowerment and area integration. The DAR is assigned as the lead agency for CARP implementation. Support services are shared with the Department of Agriculture (DA), the Department of Public Works and

17 Highways (DPWH) and the Department of Interior and Local Governments (DILG), in cooperation with Local Government Units (LGUs). The law also mandates the creation of coordinating bodies at the national, provincial and barangay levels. Members in these bodies include representatives of landowners, farmers and beneficiaries. Accomplishments and administrative costs

Official figures on accomplishments regarding land acquisitions and distribution are shown in Table 1a. Not surprisingly, redistribution is most successful for governmentowned and public lands as well as land under voluntary sale or transfer. The exception is public alienable and disposable land although the Integrated Social Forestry program (under the DENR) performs well. Least successful is land under compulsory acquisition, which can be readily attributed to landowners’ resistance. Such lands represent a fifth of CARP’s coverage. Contrary to popular impression, the worst record for compulsory acquisition is held not by those owning the largest areas under CARP coverage but rather, by those belonging to the smallest land category.

18

Table 1a. Accomplishments of land acquisition and distribution, by land type (1972-December 1999). Land type

Scope (% of total)

Distributed (% of Scope)

7

87

8

100*

19

10

6 4 9

16 2 3

37

53

16

100*

31

46

Integrated Social Forestry Areas

16

90

Total Total land (8,061,764 ha.)

63 100

73 65

Private Land Tenanted rice and corn (P.D. 27) Voluntary sale or transfer Compulsory acquisition By size category (as of 1996) 5 - 24 ha. 24 - 50 ha. over 50 ha. Total Governmentowned and Public Land Under DAR jurisdiction** Public alienable and disposable land

*

Actual distribution in excess of coverage Includes land owned by government financial institutions, KKK lands, settlements, and landed estates **

SOURCE: PARC Secretariat

An oft-cited cause of delay in the implementation of LAD is the disagreement with landowners over land valuation. Land Bank’s formula is based on a weighted average of

19 the price from comparable sales, capitalized net income and market value based on tax declaration. Unfortunately, reliable information on capitalized net income and comparable sales are usually unavailable. In the case of sales figures, CARP regulations are part of the reason for data unavailability (Bravo and Pantoja 1998). Low assessment values also preclude reliance on the tax declaration. In the absence of clear information on land values, fiscal constraints probably lead to the landowner's systematic undervaluation of their compensation. Adriano (1994), meanwhile, notes that measures undertaken to prevent corruption (to which land redistribution programs are especially vulnerable) have created a multilayered, horizontally coordinated system, further slowing down implementation. Another problem with LAD is its tendency to perform poorly in regions where land is more inequitably distributed (Table 1b). Consider the two worst performers in LAD, Regions V and VI. Based on the 1991 Census of Agriculture, the Gini ratio of landholding inequality in these regions is 0.81, compared to the national average of 0.57. These two regions account for 23.6 percent of the national CARP coverage. Over-all, it seems that land reform has managed to redistribute land only in places where it was relatively better distributed from the start (World Bank 1999).

[insert heading for Table 1b]

20

CAR CARAGA Central Luzon Ilocos Region Eastern Visayas Central Mindanao Southern Tagalog Central Visayas Western Visayas Bicol Region Total lands (8,061,764 ha.)

1.8 4.7 9.2 3.3 9.6 13.3 9 3.9 13 10.6 100

89 85 84 80 68 66 63 53 45 42 65

SOURCE: PARC Secretariat

Meanwhile, Table 2 details CARP accomplishment by support service type. Credit provision (mostly channeled through the Land Bank) to ARBs reaches a sizeable number of farmers, while total releases average nearly P7 billion annually in nominal terms. Infrastructure provision is unimpressive while no quantifiable benefits from farmer training are available. Finally, it is unclear whether CARP made a substantial difference in the quantity of services being delivered (given that these services are standard fare of government development programs) or in the targeting of these services (i.e., specifically toward new land awardees).5

5

At least for DAR-led programs, emphasis would have been toward ARCs, although at the household level

many non-ARBs may have benefited from CARP support services.

21 Table 2. Accomplishments of CARP support services (1987-July 1999). Support Service Component Amount Credit Loans released (P million) 82,290.50 Number of small farmers 6,153,380 benefited Infrastructure Completed roads (km) 5,639 Communal irrigation service 67,380 (area, ha.) Extension Number of farmers trained

2,767,348

Number of farmers provided 1,739,457 technical assistance SOURCE: PARC Secretariat

Implementation of the CARP is financed by a special fund,6 whose breakdown is shown in Table 3. These are again nominal figures spanning a 12-year interval. Total cost averages less than P5 billion a year, which is only one-fifth of average public agriculture and natural resource expenditures per annum from 1989-1998 (based on data from David and Inocencio 2000). Landowner compensation takes up less than a third of total cost, which is expected given the low rates of accomplishment in distributing private lands. Nearly a quarter of administrative cost is taken up by personnel services, in contrast to the 12 percent allocation for infrastructure, a highly capital-intensive activity. The

6

Referred to as the Agrarian Reform Fund, this fund was created by Executive Order and incorporated into

the CARL, which designates the following as the fund sources: proceeds of privatization, recovery of illgotten wealth, disposition of government property abroad, foreign funds specially designated for the CARP and government funds not otherwise appropriated.

22 administrative demands of the program under the over-all fiscal bind can explain this seemingly lopsided allocation. Table 3. Administrative costs of implementing CARP (19871999). Item Landowner compensation Other activities Credit (LBP) Infrastructure Extension Others Personnel services Other items Total

Amount (P million) 15,685 5,690.70 2,768 6,707.10 2,056.90 3,583.10 13,465.50 5,362.20 55,318.50

% of total 28.4 10.3 5 12.1 3.7 6.4 24.4 9.7 100

SOURCE: PARC Secretariat

Field evidence on CARP implementation

Official statistics mentioned earlier present accomplishments on a highly aggregated basis. This section presents findings on the progress of CARP implementation at the farm level. This discussion starts off by identifying and describing several "major data sets" specifically designed for CARP evaluation. Most of these data sets are the main vehicle for this study’s agrarian reform impact assessment. Except for the panel study conducted by the World Bank (Deininger et. al. 1999), which is based on village surveys, the major data sets emanate from the following nationwide surveys: (1) The Benchmark Survey (covering crop year 1989-1990) (2) The ARB Performance Monitoring and Evaluation System (PMES)

23 •

Phase 2 (crop year 1994-1995)

•

Phase 3 (crop year 1996-1997)

(3) The Benchmark Survey of ARCs (crop year 1993-1994) (4) MODE Impact of Agrarian Reform Survey (crop year 1996-1997)

Except for the fourth, which was undertaken by a nongovernmental organization, MODE Inc., these surveys were commissioned by the DAR to the Institute of Agrarian and Rurban Studies (IARDS), formerly the Institute of Agrarian Studies (IASt) at the University of the Philippines in Los Baños. These surveys have different sampling frames (i.e., even PMES 3 differs from PMES 2); the closest to a time-series survey is the Benchmark Survey-MODE Survey comparison, although even here significant noncomparables should be noted. Further details regarding these data sets are available in theAppendix.

Accomplishment of agrarian reform based on the nationwide studies

The PMES Surveys describe the average ARB as similar to the typical farmer. The head of the household is usually male, married and in his 50s, with three to four children, and has received little more than primary education. Other surveys with a broader class of respondents do not deviate much from this characterization. The portrait of the beneficiary as a smallholder is justified: according to the PMES 2, average farm size of the ARB is only 2.41 hectares. The MODE Survey estimates average farm size at 2.6 hectares Similarly, the 1991 Census of Agriculture and Fisheries estimates the average farm size in the country at 2.2 hectares.

24 Table 4 shows the distribution of farmers by type of tenure. Sharecropping, an illegal arrangement, accounted for a quarter of the sample in the 1989 Benchmark. The MODE survey shows that, despite six years of CARP implementation, practically the same proportion (25%) of farmers remained as share tenants. Likewise, there is no recorded increase in the proportion of owner-cultivators.

Table 4. Distribution of respondents' parcels by tenure by survey (in %). Benchmark ARC MODE PMES PMES 1989 1993 1995 1994 1996 Share tenant 25 9 26 3 7 Leaseholder 19 8 15 10 9 Owner 6 29 11 18 10 Owner-cultivator (awaiting certificate) Landless worker

35

32

34.5

23

33

Others Total

6 8 100

22 100

5.7 14 100

38 100

35 100

Meanwhile, the PMES Surveys, along with the ARC Survey, report a low incidence of sharecropping among ARBs. All of the surveys, though, report higher proportions of amortizing owners compared to the Benchmark 1989 figure. In a separate report, the MODE Survey shows only a minimal change in the number of owners. It further states that only 8.2 percent of owner-cultivators claim to have received their land from the CARP; in contrast, 58 percent inherited their land while 32 percent purchased it outright. Note that the rice and corn are the major crops raised by ARBs (57%, according to the PMES 3). Rice and corn are the main crops of all respondents in the ARC survey while rice is planted by over 60 percent of respondents in the Benchmark and MODE surveys.

25 The bulk of the reported LAD accomplishments among rice and corn farmers may actually be due to the implementation of PD 27, even prior to CARP. The data in Table 5 regarding support services is not very encouraging. The PMES reports that even among the ARBs, a substantial minority (33%) does not recall receiving assistance from any government agency or unit. Naturally, in the non-PMES surveys that included non-ARB respondents, the proportions of those receiving assistance is much lower than those reported in the PMES. The PMES 3 Survey is more upbeat: 63 percent of ARBs recall receiving government assistance. In this survey, assistance is classified by activity, namely: training (39%), fertilizer dispersal (9%), animal dispersal (7%), seed dispersal (11%), infrastructure (27%), technology transfer (31%), postharvest facilities (4%), marketing assistance (1%) and crop insurance (2%). Table 5. Distribution of respondents by availment of support services (in %, multiple responses). Availment With assistance Source of assistance DA DAR DENR DPWH LGU Others

PMES 1994 67

ARC 1993 42

MODE 1996 -

67 33 31 11 -

24.9 49 0.8 0.3 10 42.6

41.6 28.3 9.5 -

Agrarian reform accomplishment based on the panel study

The panel study observes that, consistent with the findings of nationwide surveys, a decade of CARP implementation failed to eliminate share tenancy even in rice farms. Only 12 percent of the sample in 1985 are share tenants, which was the result of aggressive implementation of PD 27 before CARP; in 1988 share tenants still account for 8 percent of the sample.

26 A more alarming trend is the simultaneous increase in landlessness in the surveyed villages. Owners comprised a quarter of the sample in 1985; 38 percent in 1998. Meanwhile, an additional 5 percent of the sample became landless. This is probably still an underestimate figure, given the possible out-migration activities. Movement up the "agricultural ladder" therefore seems to have slowed down during the CARP period. While there is an increasing share of owner-cultivators (i.e. more and more families reaching the top of the ladder) there is also an increasing landlessness (i.e., more and more families failing to get on the first rung). Former share tenants and landless workers comprised 82 percent of the new owners in 1988 (compared to 1972), but accounted for only 20 percent of the new owners in 1998 (compared to 1985). A probit estimation confirms the observation. This evidence suggests that while PD 27 had been targeted for the poor, the CARP’s aimed at other land reform objectives.. Apparently, DAR focused on completing the granting of land ownership to leaseholders and awardees of the Certificate of Land Transfer under PD 27. The success of the PD 27 program, in contrast to the laggard CARP, has been largely attributed to the coincident onset of the Green Revolution (Otsuka 1991). Effectively, PD 27 maintained the land incomes and values of the landowners at the pre-reform levels, defusing their opposition. Leaseholders and land awardees, meanwhile, received the income gains from the introduction of modern varieties. There is evidence to suggest that payments of leaseholders are only half that of share tenants precisely because of these lease controls (Otsuka, Cordova and David 1992).

Hypotheses regarding impacts of reform

An earlier section of this study hypothesized that asset inequality is negatively related to future growth, hence measures intended to reduce inequality such as agrarian reform may promote growth later. For the Philippines, however, no study regarding the link between asset inequality and growth has yet been conducted. Balisacan (1999) investigates the

27 link between initial landholding inequality and subsequent regional poverty and finds a strong positive relationship. Quite possibly the causation runs from equality to growth to poverty reduction, although further study is needed to finalize this claim. Another possibility is wealth equality’s impact on promoting local consumption linkages, which consequently expands rural industries. Only one study (Ranis, Stewart and Reyes 1989) explores the inequality-rural industry link for the Philippines but even this fails to assemble evidence from village data. Hence, the inequality-rural industry link remains an unsubstantiated, although interesting, conjecture. Focus is therefore shifted on the more specific hypotheses regarding CARP impacts.

Expected CARP impacts According to the framework, the land distribution component of the program may enhance economic efficiency. In the case of public land transfer, and perhaps even for government-owned lands, beneficiaries may in most instances be the current occupants. Hence, the transfer essentially formalizes land rights, thus, establishing tenurial security. Given that the bulk of CARP accomplishment in LAD falls under this category (63%, excluding forestry areas which account for another 16%), this is expected to be currently the major source of CARP redistribution benefits. However, note that no studies have been conducted on the productivity gained by awardees of government and public property as an offshoot of the tenure formalization under CARP. This study will hence redirect focus on the impacts of CARP on private nongovernment lands. Credit services supporting CARP also serve to overcome market failures. Meanwhile, infrastructure, technical assistance and even community organizations may be means to give farmers access to public or quasi-public goods, which their markets are ill-equipped to supply. These complementary services are perhaps essential to breaking the dynamic pattern of inequitable land distribution and its accompanying efficiency losses.

28 On the other hand, the program does undermine efficiency in other ways. Some of its features introduce distortions in agrarian markets. Tenancy regulation induces owner cultivation with hired labor, which may be less efficient. On the other hand, the permitted tenancy contract (the leasehold) exposes the tenant to more risk. The regulation of land markets further suppresses their development for land rights. Even as the program aims at expanding an ownership base, non-owners’ access to land is restricted. These regulations can further fragment the fragile credit market. In the informal sector, the shift to owner-cultivation and leasehold may sever landlord-tenant ties that are characterized by interlinked credit and cost-sharing. Restrictions on land access may also constrain the credit ties with other lenders (i.e. traders, input dealers, etc.) who typically practice informal lending where the farmer's share in the produce is used as collateral. In the formal sector, lenders such as banks may experience a dramatic climb in the transaction cost of agricultural lending, as a burgeoning mass of owner-cultivators apply for small loans. In addition, and perhaps most importantly, landowners may find themselves rationed even more severely in the credit market, as the collateral value of their land vanishes in the face of restrictions on land transactions. The protracted and unpredictable implementation of the program is another source of distortion. CARP weakens incentives to invest in land improvements during its implementation phase. Moreover, given that only agricultural lands are covered, a common criticism of the Program (aired especially by vocal farmer groups) is its failure to address evasive land conversion. In sum, these hypotheses summarize the foregoing discussions: (1) Credit, investment, and CARP beneficiaries. Land awardees obtain more credit and accumulate more assets without the program. (2) Land access and CARP. The program restricts access to land, thereby suppressing upward mobility of the rural poor. (3) Credit, investment and landowners

29 •

The program reduces landowner's investments in land improvements.

•

The program reduces aggregate investments due to the diminished collateral value of agricultural landand to scale diseconomies in smallholder lending.

(4) Land use and CARP. The program introduces inefficient land use practices due to uncertainty in land rights and excessive regulation of land markets.

CARP impacts on landowners and land use

The incentive effects of CARP on land investment are uncertain. A survey made by of the Management Association of the Philippines in 1990 found that over 60 percent of 39 respondents (farming over 72,000 ha.) either reduced their investments or shelved expansion plans (Llanto and Estanislao 1993). Even granting that this survey is credible, the only other evidence remains anecdotal. Meanwhile, the erosion in the collateral value of land has been looked into by Llanto and Dingcong (1994). They gathered data on borrowings by landowners and agrarian reform beneficiaries and using logit regression, concludes that the probability of being rationed does not depend on the size of agricultural land, which suggests that such property has lost its collateral value. However, the magnitude of the implied credit reduction is unclear. Clarete (1992, as cited in Llanto and Estanislao 1993) has attempted to quantify the welfare losses arising from the loss of collateral value of land, using Computable General Equilibrium (CGE) methods. He estimates the productivity of farming and other primary sectors to fall by 4.7 percent and 3.7 percent respectively as a result of this effect. The annual decline comes to around P2 billion a year. These similations are indicative of the orders of magnitude associated with these losses.. Some evidence on rising transaction cost has been gathered by Casuga (1994). Based on her sample of 64 formal sector creditors, she measured the transaction cost in 1986 (before CARP) as 4 centavos per peso loan for rural banks. Three years later (after one

30 year of CARP implementation), the transaction cost grew to 9 cents per peso. However, she rightly cautioned against attributing this increase to the program. Likewise, little can be said about the actual land use distortions inadvertently imposed by CARP. Land conversion, a hotly disputed issue in land use, does not appear to reach alarming proportions; approved conversions totaled only 1.2 percent of DAR coverage in 1997, and only 1.3 percent of total rice areas. Even taking into account illegal conversions, such behavior cannot on the whole be seen as endangering CARP objectives. Nevertheless, the future potential for expanded land conversion is a cause of concern, as 88 percent of municipalities lack a land use plan (Gordoncillo et al. 1998). In sum, there is a mild confirmation of hypotheses 3 and 4 but no hard evidence regarding the gravity of the hypothesized welfare losses.

CARP impacts on actual and potential beneficiaries

To determine the impacts of CARP on actual and potential beneficiaries, this section examines issues of credit, capital accumulation and land access. As a prelude, the following reviews existing data on productivity and incomes.

Production and earnings

31 Yield comparisons by crop are presented in Table 6. The ARBs produced approximately the same yields as the average farmer. The only serious difference is the estimate for rice yield from the MODE Survey. Smaller surveys also suggest that yields Table 6. Crop yields of respondents by type of crop by survey (in tons/ha). Crop

National average 1994 1996 2.9 2.8 1.8 1.5 1.2 1.3

PMES 1994 2.9 1.6 1.2

PMES 1996 2.7 1.7 1.4

MODE 1996 3.7 1.7 -

Rice Corn Copra (annual) of CARP-affected farmers are similar to national yields, e.g., Geron (1994).

Table 7 presents the magnitude and sources of income. Poverty incidence of families among ARBs is over 60 percent. According to the MODE Survey, around 70 percent of respondents were poor (using a slightly higher poverty line than that employed in the PMES). Note though that, except for the PMES 3, surveys compute sample poverty using total household incomes, hence, requiring the imputation of a family poverty line (unadjusted for household size). These figures are higher than official figures on rural poverty of households (44.4% in 1997). The appropriate data that can be compared with that on ARBs are the poverty rates in agricultural households, for which no figures are available in 1997. Previous studies suggest that agricultural poverty is higher than rural poverty.7

7

If Balisacan's (1997) calculations are extrapolated to estimate the divergence between national poverty

and poverty among agricultural-dependent households, the resultant figure is closer (58.5%) to rural poverty of households.

32 Table 7. Income, poverty, and income sources of respondents by survey. ARC 1993

Average income 51,939 (pesos) Proportion below 54.5 line(%) Sources of income (%) Farm 46.6 Off-farm 8.9 Nonfarm 44.5

MODE 1996 -

PMES 1994 47,884

PMES 1996 54,631

71.7

~ 60

62.8

43 4.9 52.3

47.9 5.5 46.6

42.8 2.7 54.5

Income from farming accounts for less than half of total household income---a fact corroborated by three surveys conducted in the mid-1990s. Meanwhile, off-farm incomes accounts for only 5 percent to 10 percent of household income; nonfarm income accounts for a large share in earnings. The surveys covering 1996 show that over half of family income is from nonfarm sources. In the ARC Benchmark Survey, nonfarm incomes were apparently the source of inequalities as some of these are OCW remittances; in the PMES 2, however, the distribution of the share of nonfarm income in total income did not vary much across income deciles (Bravo and Pantoja 1999). Household income figures, however, may not be fully comparable across surveys. Consider the net farm income per hectares (gross of rent): the MODE Survey has the figures P17,942 for rice, P6,892 for corn, and P3,830 for coconut. Contrast this with the corresponding PMES 3 estimates: 35,718 pesos, 12,801 pesos, and 2,508 pesos, respectively. The divergences are perhaps too large to be accounted for simply by the profit edge of ARBs.

33

Credit

According to the MODE survey, 47 percent of respondents are dissavings. The PMES 3 reports that 54 percent of ARB respondents made borrowings, a figure close to the 56 percent of the MODE Survey. According to the latter, a greater proportion of ARBs and leaseholders are borrowers, compared to other respondents. Table 8 details the sources of credit. The PMES Surveys seem to show a large increase in credit access. This is due to the unusually low proportion of respondents (30%) who claimed to have borrowed in 1994. Table 8. Borrowing sources of respondents as % of all respondents, by survey (multiple responses). Source Formal sources Cooperative Other formal sources Informal sources Traders Input dealers Moneylenders Relatives and friends Others

PMES 1994 9.1 2.2

PMES 1996 13.5 7.6 5.9

MODE 1996 6.4 3.4

21.5 6.7 2.5 6.9 5.3

40.5 13.5 15.7 6.5 15.7

21 1.1 5.5 11.4

0

2.7

6.7

6.9

Borrowing from formal sources appears quite limited. Instead, respondents who borrowed relied mostly on informal credit. Contrary to popular impression, moneylenders are not the main source of informal credit. Surveys show an important role played by credit coming from buyers (MODE) and suppliers (PMES 3). Credit sourced from buyers

34 is probably an interlinked scheme while the credit from suppliers is probably trade credits from fertilizer, pesticide and seed dealers. Informal lenders demanded collateral only from a few of the borrowers (27%, according to the MODE Survey). The PMES surveys contain data on ARB amortization compliance. A little more than a tenth of ARBs were irregularly paying amortization in 1994. By 1996, this proportion reached one-fifth of ARBs. While the survey rules out ARB default on a massive scale, no information is available regarding the severity of the default threat among irregular payers. Nor have any studies been conducted on the measures taken by Land Bank to deal with these cases, particularly with ARBs who ceased repayment (7% in 1994). These data show that ARBs are no less productive than average small farmers and even slightly poorer than the typical agriculture-dependent household. There is a significant dependence on nonagricultural earnings, and there are indications that such dependence is growing. Many ARBs borrow, but mostly from the informal sector, where fixed asset collateral is seldom required. Note that the agrarian reform regulations preclude even full-fledged owners-beneficiaries from using the awarded land as collateral for at least ten years upon the awarding. Nor can government credit be expected to substantially offset this, given weak availment rates of cooperative credit, which is the main channel of such credit support. The major question, of course, is whether CARP or agrarian reform in general positively affected productivity and income. The question cannot be answered using the available nationwide studies for two reasons: First, the issue of causation, which can be most effectively addressed by regression analysis,8 is not covered by any nationwide studies.

8

Of course, regression analysis does not establish causation per se but can substantiate the magnitude of an

assumed causation. Justification of the assumed causation requires appeal to principles or evidence other than the regression analysis itself.

35 Second, sampling frames differ across surveys, making inter-temporal comparison difficult, if not impossible.9

Capital accumulation

The panel study yields productivity and income trends that suggest improving households’ circumstances due to agrarian reform. Average household incomes in real terms rose by 46 percent between survey rounds; by 1998, average household expenditure was above the World Bank poverty line. There is, however, no trend found in terms of more diversity of income sources. Agriculture continues to provide the bulk of household earnings. This study used multivariate regression analysis to isolate the contribution of agrarian reform to the trends. The method consists of regressing (in differenced form) an outcome variable against household attributes, time trend and a reform beneficiary dummy. The study found that ARBs were able to provide education to their children as well as

accumulate more assets. That is, agrarian reform increased investments in human and other capital. The magnitude of the welfare gains is significant by any standard, as land reform beneficiaries earn 30 percent more than the national per-capita income. This study explains these wealth effects based on the idea that an asset transfer lends to growth by encouraging investments and easing credit constraints. The study, however, gives no direct evidence regarding agrarian reform’s impact on credit.

9

An example of a pitfall in the use of these surveys can be seen in the report of Garilao (1998), which

alleges that ARB income increased, by comparing average income of farmers from the Benchmark Survey (P47,884) and that of the ARBs from the PMES 2 Survey (P56,646). The comparison is obviously faulty given that only nominal incomes have been measured, and that any number of other factors may have led to an income change. In addition, it is difficult to compare households composed of affected farmers in a broad sense (Benchmark Survey) with ARBs in the narrow sense (PMES Survey).

36

Land access

Nationwide studies are largely silent about land access. Village surveys over the last two decades have on the other hand substantiated the notion that agrarian reform restricts access to land. The direct evidence takes the form of tenant eviction: this may have befallen one-fifth to one-half of tenants in Central Luzon and Panay, going by the findings of a survey of five villages in conducted in 1986 (Otsuka, 1991). The emergence of permanent labor contracts is a persuasive circumstantial evidence regarding the decline of land access opportunities. A new kind of labor contract in Central Luzon, called the kasugpong, is probably substituting for the prohibited tenancy contracts (Hayami and Otsuka 1993). The proportion of permanent laborers versus landless workers in some Central Luzon villages has increased from virtually zero to around 30 percent by the 1980s, after two decades of PD 27 implementation. Under this permanent labor contract, a worker receives a fixed amount of paddy rice or a small proportion of harvest in exchange for services throughout the cropping season. However, by the fixed nature of the payment to these workers, labor requires constant supervision and is prone to shirking. In the Central Luzon farms, the residual profit from kasugpong farms is around a quarter below those of family-operated farms. More direct evidence on the reform-land access link is given by the panel study. A nonparametric regression finds that ownership of land is strongly related with size of landholding, whereas under a healthy rental market no such relationship could be found. A parametric regression confirms that area cultivated is strongly affected by size of owned land. The welfare impact of restricted access could be substantial: the consumption of the landless who later successfully gained land access was 30 percent greater than those who failed. On the whole, the negative impact may well have exceeded the benefits of agrarian reform. Prohibitions on share tenancy as well as rent controls deny the landless the

37 opportunity to rise in the "agricultural ladder." Adverse equity and efficiency consequences are expected to the extent that poverty is concentrated in this sector, as is the finding of income studies (Balisacan 1993; David and Otsuka 1994).

Implications for policy

If the hypotheses mentioned earlier hold for the Philippines, then the policy prescriptions for changing course in agrarian reform become clear. The experience of World Banksupported land reform programs may be used to enumerate a set of broadly stated best practices (Deininger and Binswanger 1999). The World Bank policy agenda proposed in the mid-1970s were focused on the following desirable goals: owner-operation, active land markets and egalitarian asset distribution. Their merit has been largely confirmed over a quarter of a century. Add to this a few more hindsight/ideas. One is that land transfers should be accompanied by broader efforts to improve related markets and infrastructure. Another is that the reform process should take into account the considerable potential of markets for land rentals and sales in promoting land access and egalitarian land distribution. Restrictions on tenancy and land markets must therefore be lifted (Otsuka 1999). A recommendation is to shift away from crudely coercing land distribution and toward the more subtle form of coercion, land taxes. Progressive taxes on land, and taxes on idle land are highly favored as relatively efficient instruments for indirect land redistribution (Hayami, Quisumbing and Adriano 1991). The revenue potential from progressive land taxes is respectable, judging from the CGE simulations of Habito (1989). The administrative demands of levying such taxes are however daunting, particularly at the national level. First, landowners would be likely to misdeclare land ownership and quality. Second, land taxes encounter considerable coordinated opposition from landed interests, particularly when they see it as an pure tax burden, whereas the constituency for such taxes are poorly organized (Skinner 1993).

38 An emerging consensus on land reform is to back up tax instruments with "negotiated" solutions and implement in a decentralized fashion with reliance on beneficiary participation (Deininger 1999). This strategy is counter to the "voluntary" nature prevailing in the CARP. Deininger et al. (1999) suggests that the opposition may be mitigated by tying revisions in the tax regime with the lifting of restrictions on land markets or even of ownership ceilings. While other analysts refrain from recommending a repeal of retention limits ( as advocated by Hayami, Quisumbing and Adriano), there is a broad consensus among economic analysts regarding the “reform of land reform.” Research issues - Agrarian

This summary recapitulates the general link between equity and growth in the Philippines as well as the impacts of agrarian reform on landowner investment and intended beneficiaries. First, the positive relationship between land ownership equality and future growth

is yet to be convincingly established from Philippine experience. It is a fact, however, that asset equality is positively related to future poverty reduction. Because of this and other considerations, there are good reasons to expect the posited relationship to hold. The requisite test is not too demanding; one only needs to conduct an analysis, perhaps on a regional level, and compare area growth rates with some measure of initial asset inequality. The available measure pertains to operational landholdings (available from the Census of Agriculture and Fisheries). A more demanding but still feasible exercise may be to use as a measure the inequality of land ownership. It is possible to work on the existing DAR and Land Registration Authority records to assemble a database on recent land ownership patterns by region, and compare this with subsequent growth rates. In this vein, much work still needs to be done on the relationship between income inequality, local linkages and the growth of rural-based industries. Data from ARCs can help frame this particular thrust.

39 Second, evidence concerning foregone landowner investment due to agrarian reform

remains little more than anecdotal. An intensive data gathering---from landowners, banks to other financial intermediaries---is recommended to quantify the investment losses attributed to agrarian reform. Preliminary efforts based on rapid appraisal methods will be a good start. Third, there are yet no nationally valid evidence on the contribution of agrarian reform

to incomes and assets of beneficiaries of private land redistribution. Data gathering should focus not merely on ARBs but rather on representative samples of the rural population; data should focus on production, income, credit, assets and education. If possible, impacts from various agrarian reform interventions should be isolated. Moreover, nationwide surveys of rural households should be undertaken repeatedly and consistently. The past decade has seen several lost opportunities for a time-series study that tracks trends for beneficiaries and non-beneficiaries. Fourth, the implications of land market regulation on land access, equity and poverty

should be quantified at the national level. Findings on land access trends based on village studies need to be integrated at the national level; this is a must since out-migration may conceal the magnitude of rental market suppression. Hence, these nationwide studies must be representative enough to include a sufficient proportion and number of the landless. The existing nationwide surveys nevertheless can be mined for further information, if not about trends, at least about the link between land access and ownership. Moreover, a well-rounded analysis should include those on agrarian contracts that may arise to circumvent the land market restrictions. Research should proceed beyond the already well-researched villages and provinces and into other informal rural markets all over the country. Fifth, practically no work has been done regarding the productivity and welfare impacts

of providing tenurial security to occupants of government and public lands. The gap in agrarian reform research is striking, given the concentration of land distribution accomplishment in government and public lands only. Research specifically intended at

40 quantifying CARP impacts only (i.e., not the cumulative effects of all land programs undertaken in the country) should probably take this issue as a starting point. Sixth, policy research may begin to seriously investigate tax-oriented, and otherwise

noncoercive schemes for agrarian reform. While evidence on the harmful effects of current land reform efforts is yet to be finalized, the implementation roadblocks currently being encountered pose strong objections to the drive to complete CARP within its current set-up. Innovative approaches to rural land markets should now begin to draw the attention of policy analysts and policymakers.

Property Rights Reforms In Fishery And Forestry

There are three reasons why the current fishery and forestry use tends to lead to excessive extraction. The first two are static allocation problems, while the third refers to dynamic tradeoffs between current and future uses of a resource. The following incorporates standard textbook treatment of these issues (e.g., Johansen and Lofgren 1985).

Open access

Fishery and forestry resources are regarded as common pool resources (CPRs) that are subject to multiple use, difficulty in exclusion and rivalry in the extracted resource, i.e., fish catch or logs (Ostrom, Gardner and Walker 1994). The problem of the "commons" (i.e., an open access problem) was pointed out in a seminal paper by Hardin (1968). Consider a resource stock that yields a homogenous harvest to users upon application of effort. Suppose the cost of each unit of effort and the price of the harvest are constant. Also assume the set of users is the industry, and industry effort is subject to diminishing

41 marginal product. In the region of diminishing returns at the margin, average product exceeds marginal product and is falling. Open access implies free entry; hence, equilibrium is reached when the marginal user earns zero profit. This implies equality between average product and marginal cost of effort. The optimum though is at the point where marginal product equals marginal cost. The equilibrium effort implies over-harvesting of the resource. A user is unable to take into account the external effect of his/her own effort on the harvest of others.

Offsite services

Natural resources such as forests provide a wide range of environmental services (watershed maintenance, erosion control, wildlife preservation, etc.). When forests are felled, the loss of these services involves a cost above that of the extractive activity itself (e.g., timber cutting and hauling). The extraction cost and offsite cost together comprise the social cost. A private sector firm will extract timber until the marginal cost of extraction equals the unit value of the timber. Since social optimum involves equality of the unit value of the timber with the marginal social cost, the private firm's equilibrium involves excessive extraction. Take for instance a logging company’s decision that does not consider the costs imposed by downstream siltation of lakes and reservoirs; here, too little forest cover in the end will be maintained.

Intergenerational concerns

The foregoing discussions refer to externalities imposed by a user on other current users. There is still a different type of externality imputed to future users. This is the concept of

sustainability. Consider once again a given resource that yields a homogenous harvest. To isolate the intergenerational aspect of the exploitation problem, consider the case of a privately-owned resource, where extraction involves no offsite costs. Let the discount rate be r. Various stock levels of the resource are possible, ranging from the minimum threshold to the maximum carrying capacity. Below the threshold, the

42 population declines to zero; beyond the carrying capacity, the stock will also tend to decline. In between the threshold and the carrying capacity, the stock grows. The growth rate at first increases with the stock level, then decreases to zero as it approaches the carrying capacity. At equilibrium, the harvest equals the growth rate; hence, the resource stock remains constant. The equilibrium stock is that level where the (positive) change in growth rate equals the discount rate. If the discount rate increases, the equilibrium stock falls (so as to realize greater changes in the growth rate). It is quite possible for the discount rate to be sufficiently high such that the short-term equilibrium stock is set below the threshold level---implying complete extraction of the resource in the long run. The problem with the equilibrium solution is that the value of the resource and the discount rate are set only by the current generation. Compared to valuation that represents scenarios for future generations of users, the harvest price may be too low or the discount rate too high. The problem becomes more stark when long-run equilibrium entails extinction of a species; the willingness to pay of future generations for some positive population is not captured by the market.

Regimes for reducing exploitation

Regulation and taxation instruments. To correct these externalities, the country should find a cost-effective means of limiting extraction. The most direct method is command and control. However the geographic sprawl of the CPRs as well as the transportation costs have, in many countries, rendered this instrument largely ineffective (Hyde et al. 1996). Another means is to impose harvest fees or taxes. If set at the right rates, taxes and fees can lead to a complete internalization of costs, although their implementation costs may be high or prohibitive. Private (individual) land rights. A recent alternative policy is the promotion of property regimes in CPRs. One option is to confer use rights to individuals. Such an option is

43 feasible for forestland, where numerous titling or other tenure programs have been undertaken in developing countries. (It should be noted that, even in theory, conferment of use rights can be effective only against the open access problem. It cannot be regarded as a means to adequately account for offsite externalities.) In practice, establishing private property arrangements must contend with the costs of enforcement. The evolutionary theory predicts that, where benefits of internalization exceed costs, private property rights will evolve. The absence of such conditions is prima

facie evidence that private property arrangements are too costly. Forest products have low value as well as low marginal extraction cost, whereas the cost of limiting access to ___________ is typically large, even for a private landholders (Hyde et al. 1996). Tenurial programs are often coupled with agro forestry projects. In the framework of the evolutionary theory, this is to raise "the benefits of internalization." Internalization of costs can be facilitated by extension efforts and subsidies packaged as conservation programs. So will a titling program. Successful cost internalization projects suggest that profitability is a critical element; quicker gestation projects are favored by farmers as well. Thus, tenurial security promotes investment and innovation. Interestingly, the greatest threat to security comes from government regulation (Current et al. 1995). For instance, logging restrictions can effectively nullify tenurial arrangements and reduce the incentive to protect the property rights. Ironically, even if private property rights can be effectively enforced, land titling programs (or even the expectation of such programs) may actually hasten deforestation. The perverse possibility arises because occupancy is usually the criterion for obtaining a title---hence, providing an incentive to undertake clearing activities (Angelsen 1999). Common property. The third option for CPR management is to promote common (in contrast to private) property arrangements. The durability of common property arrangements, some of them rooted in ancient tradition, has motivated numerous attempts to analyze their effectiveness in resource management. At the simplest level, communal

44 arrangements realize economies of scale in protection activities. In terms of production, however, individual rights are typically bestowed on the basis of land clearing and occupancy. Hence, customary tenure is not exactly inimical to commercial activities such as agro forestry (Otsuka 1998). Collective production is not a common feature of communal arrangements, contrary to popular belief; rather, such arrangements provide public goods, enhance equity, undertake risk-reduction or help break seasonal labor bottlenecks (Deininger and Feder 1998). Factors conducive for a group to govern its members effectively have been compiled in a famous list by Ostrom (1994), which is based on a wide range of case studies. These factors are: (1) exclusion of nonmembers from the resource; (2) appropriateness to local conditions; (3) membership participation in rule setting; (4) accountability in monitoring; (5) application of graduated sanctions; (6) presence of low-cost conflict resolution mechanisms; (7) recognition by the formal authority; and (8) nesting of cooperative groups in large organizations.

Meanwhile a formal approach typically employs game theory. Common property arrangements are interpreted as a cooperative equilibrium achieved by the players. The more appealing models employ repeated interaction, where the central idea is that exclusion from future benefits can lead to noncooperative behavior. For cooperative equilibrium to be possible, the player must perceive the number of interactions to be indefinite, retaliation from other members credible, and the future sufficiently important (Seabright 1993). Unfortunately, as expressed in the Folk Theorem, cooperative behavior is not the only possible equilibrium. Nor is it clear that a cooperative equilibrium is robust to environmental and group change.

45 One extension of the theory of cooperation is to examine the role of history in the evolution of social norms. Sethi and Samanathan (1996) describes the development of social norms as the evolution of strategies in terms of "replicator dynamics". Strategies are increasingly adopted when their payoffs yield more than the average payoffs. They find that norms of restraint and punishment can be stable, even against the entry of narrowly self-interested players. Another related theory recognizes that communal arrangements need not fit into a cooperative/noncooperative dichotomy; rather, a wide range of success in various aspects of resource management are possible. McCarthy, de Janvry and Sadoulet (1998) regard cooperation as a matter of degree and subject to variable costs. Their model provides a flexible framework for identifying factors that raise or reduce the equilibrium degree of cooperation.

Programs and reforms in the forestry and fishery sectors

The DENR takes the lead in administering programs and implementing policies for the natural resource sector. For forestry, extraction is regulated under Timber Licensing Agreements. The agency is also implementing several programs to address forest denudation as well as upland poverty. the CARP, too, already has tenure programs in forest land, referred to as the Integrated Social Forestry Program. Beneficiaries of this program are issued nontransferable Certificates of Stewardship to provide them security of tenure. The oversight of the projects has been mostly devolved to the LGUs since 1991. The Community Forestry Program, meanwhile, assigns forest protection and management to organized communities. The privilege to use and sell forest products is formalized in a 25-year Community Forest Management Agreement. By 1997, there were 66 such agreements covering 173,298 hectares of land. Other community-based programs

46 are the Forest Land Management Program (covering reforested areas) and the Regional Resources Management Program. For fisheries, legislation has been consolidated in the Philippine Fisheries Code of 1998. The other important laws are the Local Government Code (LGC) of 1991, and the Agriculture and Fisheries Modernization Act (AFMA) of 1997. The LGC defined the scope of municipal fisheries to cover waters up to 15 kilometers from the coast. Boats above 3 tons (which are classified as commercial vessels) are not allowed to fish in these areas. It also expanded the authority of LGUs in administering these areas. The municipal authority was empowered to enforce fishery laws, license municipal fishers, grant privileges to organized fishers in constructing immobile gears and regulate local waters. Meanwhile commercial waters remain the responsibility of the national government, specifically the Bureau of Fisheries and Aquatic Resources. The Fisheries Code as well as the AFMA explicitly expressed the principle of sustainable development in the management of agricultural and fishery resource (Israel and Roque 1999). For commercial waters, licensing and fishing permits are supposed to reflect resource rents as well as regulate harvests to the maximum sustainable yield level. The code also enumerated various sanctions on illegal fishing gear, exploitation of sensitive resources, catch limits, restricted fish species, etc.

Evaluation studies for the Philippines

Incentives and environmental degradation

Forest denudation in the country is typically associated with upland migration; estimates of upland population range from one-tenth to one-third of the total population. This suggests "push" factors, given declining person-land ratios and nonfarm employment opportunities in the lowlands (Cruz and Repetto 1992). Upland dwellers as well as coastal fishermen comprise some of the poorest sectors in the country, with subsistence

47 activities being the norm. Upland poverty itself is sometimes thought to be a contributor to unsustainable land use practices. This notion is, however, not empirically substantiated (Grepperud 1997). Coxhead and Rola (1998) find that "pull" factors in the form of price incentives are also important, particularly for the spread of erosive farming. Lack of access to credit constrains the adoption of conservation measures. Increasing access to credit has, however, an ambiguous impact on soil erosion as land clearing is also positively related with credit access. For the fisheries sector, Israel (1997) found that virtually open access regime in commercial waters resulted in fishing effort far in excess of that required for either maximum economic yield or for maximum sustainable yield. In 1994, fishing efforts should have been reduced by about 12 percent to attain the maximum sustainable yield or by nearly half (45%) to attain the maximum economic yield.

Environmental taxes and subsidies

Environmental observers agree that extraction charges are too low, given the rents earned. In the case of forestry, an early calculation (de los Angeles 1989) found that charges extract less than a tenth of resource rent. For fisheries, large resource rents are likewise being earned but the fees have not been adjusted since the early 1980s (Israel and Roque 1998). However, fee increases may meet strong political opposition; Elazegui and Paunlagui (1999) cite an example of a municipality which could not raise fees due to vehement local objections. Government programs to promote sustainable technology also promote adoption of conservation measures. Adoption rates among cooperators in the Central Visayas Regional Project are higher than that among noncooperators. Moreover, the degree of participation is positively related to the adoption choice. Incomes of cooperators are also

48 shown to be higher than incomes of noncooperators. The former has also increased faster over the program’s eight-year implementation (Francisco 1994).

Tenurial security and property rights

De los Angeles (1994) claims that the link between upland conservation and property rights is "no longer debatable". She cites studies which found that the extent and pace of adoption of conservation measures differ between CSC holders and non-CSC holders. Coxhead and Rola (1998) confirm that less secure tenure in the uplands is associated with the adoption of erosive farming. As mentioned earlier, tenurial security may be undermined by government regulation. The threat of imposing a total log ban casts a veil of uncertainty over upland property rights arrangements. Most likely, given the enforcement costs, elimination of formal rights will lead to informal encroachment. The de facto open access state may therefore reassert itself. Extreme logging restrictions may unintentionally promote forest destruction (de los Angeles and Oliva 1996).

Promotion of community-based management

Building community organizations is an important factor in the adoption of agro forestry schemes (Francisco 1994). The promotion of co-management (where the State retains ownership of the resource but users also undertake resource management) in fisheries is currently an active research area of the WorldFish Center (formerly the International Center for Living Aquatic Resources Management, or ICLARM). Among the Asian countries considered (Philippines, Vietnam, Thailand, Malaysia, Indonesia and Bangladesh), the Philippines was singled out as having the most experience with community-based management of coastal resources as well as the strongest set of supporting policies and laws.

49 These studies evaluated the ingredients for sustainable common property arrangements, identified by Ostrom (1990, 1994). As listed in Pomeroy, Katon and Harkes (1999), these are: (1) clearly defined boundaries; (2) clearly defined membership; (3) group cohesion; (4) existing organization; (5) positive net benefits from the member's viewpoint; (6) participation by those affected; (7) enforcement of management rules; (8) legal rights to organize; (9) community level cooperation and leadership; (10) decentralization and delegation of authority; and (11) coordination between government and community.

Among these, those of “high importance” are numbers 1, 2, 5, 6,7 and 9. To illustrate: in San Salvador and in Malalison Island in the Philippines, the marine sanctuary was clearly demarcated with buoys. All members of the fisher organization were involved in making and changing the rules. The NGOs devote much time and effort in educating fishers about the benefits and costs of co-management. Co-management was more successful in communities where fishers had positive attitudes toward collective action, and where a strong local leadership was present. Studies that quantify the benefits and costs of co-management are, however, sparse. Katon et al. (1997) analyzed beneficiary perception of the quality of life improvements in Cogtong Bay, Central Visayas. Co-management in this area began in 1989, when the national and municipal government, together with local fishers, established a regime of coastal resource management. The co-management project was found to be successful in promoting positive and statistically significant changes in its performance indicators,

50 except for the household income indicator. Not surprisingly, the most significant changes were observed for indicators related to "empowerment." As for the cost side, a study by Abdullah et al. (1997) measured the transaction costs of a fishery co-management project. This study categorized transaction cost as either information, collective decision-making or operational costs. the following are the results of the study (in pesos):

Co-management Centralized management

Years 1-2

Years 3-4

Years 5-8

1,547,272

1,113,250

1,200,200

446,895

467,542

2,830,847

Figures on centralized management are derived from key informant interviews and other studies. Its total cost is slightly lower than that of co-management (P3.86 million versus P3.74 million). However, the bulk of the transaction cost of co-management take the form of initial start-up costs. It may be conjectured that co-management more than compensates these initial costs by requiring lower outlays in later years. Enforcement costs may eventually be lower as community members are more likely to comply with rules formed under participation, than with rules that are externally imposed.

Research issues

- Fisheries and Forestry

The research issues for CPRs can be divided into three broad categories, namely: the link between property rights and environmental degradation; the transaction costs of establishing property rights; and the appropriate set of instruments, in combination with the property rights regime, for managing CPRs.

51 The link between conservation practices and tenurial security is well established. What is unclear is whether the promotion of formal individual rights under conventional programs contribute significantly to the control of environmental degradation. Of special concern is whether perverse incentives for land clearing and migration might be created by tenure programs. Except for a few studies on fishery co-management, there are no estimates available regarding the benefits as well as costs of establishing property rights regimes. More work needs to be done in this area, especially in terms of evaluating and comparing alternative regimes. Emphasis should be both on generating concrete figures over time for particular cases so as to obtain an idea about orders of magnitude as well as relating costs over time to the locality’s social and physical environment. An interesting hypothesis for research is that community-based management---when compared to centralized management--involves large initial investment but low recurring costs. Finally, research should also examine further the appropriate combination of other interventions with the promotion of property rights. Consider pricing issues: a criticism is that fees collected for logging concessions and commercial fishing licenses are too low to capture resource rents. The appropriate level and structure of fees as well the system of fee setting are yet to be specified. Further studies should therefore focus on the appropriate level of fees and process of adjustment. Also, as the Fisheries Code includes a very limited provision on market-based instruments, analysis of policies in this direction will be a valuable input to future legislation. To date, studies that focus on the increasing role of markets in the tenurial and access schemes are a controversial yet poorly-researched area. The devolution of oversight functions over the ISF Program as well as of municipal waters presents another important issue. While most observers approve the general intent of the Local Government Code, research on the benefits of the changing of hands on uplands and coastal resource management is yet to be rigorously studied. As most recommendations center on improving the capacity of local governments to undertake the

52 environmental protection, research on governance should specifically focus on capacity-building proposals. Conclusion

The agriculture and natural resource sectors in the Philippines have undergone extensive property rights reforms, particularly from the late 1980s. Recent literature on property rights is replete with empirical evidence across countries and over time, on both favorable and unfavorable efficiency impacts of such reforms. Some carry hypotheses on equity and sustainability and on links between efficiency and equity. Unfortunately in the case of Philippine agriculture, much of these hypotheses have yet to be established. Findings of past studies are mostly based on case studies, where their applicability over a bigger locality is inconclusive. Greater effort should be directed to achieving wider generality. Also, nationwide studies for agrarian reform do not permit time-series comparisons. Nor do they address large coverage areas for reform such as occupants of public and government lands. Though opportunities have clearly been missed, these are part of a rich potential for further research. The very slow reform implementation itself consequently permits extended comparisons between beneficiaries and non-beneficiaries, which is essential for any statistical analysis of causation. While the literature has indeed burgeoned over time, the absence of a concerted, systematic research aimed at gathering evidence, tracking changes over time, and testing hypothesis arising from a coherent framework is regrettable but not irreparable. By paying heed to what is known and not known about the impacts of property rights reform, further research work can bridge the gap within this decade.

53

References Abdullah, N., N. Mustapha, R.Pomeroy, E. Genio, A. Salamanca (1997). Transaction

Costs And Fisheries Co-Management. Universiti Pertanian Malaysia & ICLARM Serdang Selangor & Manila: Fisheries Co-management Research Project.

Adriano, L. (1994). “DAR, Land reform-related agencies, and the CARP: A study of government and alternative approaches to land acquisition and distribution.” Philippine Institute for Development Studies (PIDS) Discussion Paper No. 94-13.

Aghion, P., E. Caroli, and C. Penalosa (1999). “Inequality and economic growth: The perspective of the new growth theories.” Forthcoming in Journal of Economic Literature.

Alwang, J. and P. Siegel. (1999). “Labor shortages on small landholdings in Malawi: Implications for Policy Reforms.” World Development 27(8):1461-1475. Baland, J.-M. and J.-P. Platteau. (1997). "Wealth inequality and efficiency in the commons (Part 1)." Oxford Economic Papers 49(4):451-482.

Balisacan, A. (1993). "Agricultural growth, landlessness, off-farm employment, and rural poverty in the Philippines." Economic Development and Cultural Change 41:533-562.

54 Balisacan, A. (1999). “Land inequality, asset accumulation, and poverty in rural areas of the Philippines.” Paper presented at the Biennial Convention of the Philippine Agricultural Economics and Development Association, Philippine Trade and Training Center, 19 February.

Balisacan, A. (1997). “Getting the story right: Growth, redistribution, and poverty alleviation in the Philippines.” Philippine Review of Economics and Business 34(1): 1-37.

Banerjee, A. and A. Newman (1994). “Poverty, incentives, and development.” American

Economic Review 84(2):211-215.

Benjamin, D. (1995). “Can unobserved land quality explain the inverse productivity relationship?” Journal of Development Economics 46(1): 51-84.

Binswanger, H. and K. Deininger (1997). “Explaining agricultural and agrarian policies in developing countries.” World Bank Policy Research Working Paper 1765.

Binswanger, H., K. Deininger, and G. Feder. (1995). “Power, distortions, revolt, and reform in agricultural land relations.” Handbook of Development Economics. Volume 3. J. Behrman and T. Srinivasan, eds. Amsterdam: Elsevier.

Boserup, E. (1965). Conditions for Agricultural Growth: The Economics of Agrarian

Change Under Population Pressure. New York: Aldine Publishing.

55

Bravo, M. and B. Pantoja (1999). "Beyond 2000: Assessment of the economy and policy recommendations - social sector dealing with agrarian reform." PIDS Discussion Paper Series 98-09.

Briones, R. (2000). “Geographic patterns of specialization and vertical contractual relations in rural industries: Three essays on the development of the rural nonfarm economy.” Unpublished Ph.D. dissertation. Quezon City: University of the Philippines, Diliman.

Carter M., and F. Zimmerman (1998). "Can time and markets resolve the agrarian question? Microeconomic insights on the persistence and costs of land ownership inequality in Latin America." Paper prepared for the conference on Asset Distribution, Poverty and Economic Growth. Brasilia 14 –17 July.

Casuga, M. (1994). “Transaction Cost Under an Agrarian Reform Regime.” Financial

Intermediation in an Agrarian Reform Regime. Pasig City: Agricultural Credit Policy Council.

Clarete, R. (1992). “An economic analysis of selected policy reforms in the agribusiness sector: The Agribusiness Sector Assistance Program (ASAP).” Processed.

Coxhead, I. and A. Rola (1999). “Economic development, agricultural growth and environmental management: An investigation of their linkages in Lantapan, Bukidnon.”

56 Institute of Strategic Planning and Policy Studies Working Paper 98-04. University of the Philippines, Los Baños.

Cruz, W. and R. Repetto (1992). “Environmental effects of stabilization and structural adjustment programs: the Philippines case.” Washington, D.C.: World Resources Institute.

David, C. and K. Otsuka (1994), eds. Modern Rice Technology and Income Distribution

in Asia. Manila: International Rice Research Institute.

David, C., and A. Inocencio (2000). "key indicators for public expenditure in agriculture, national resources, and the environment" Processed.

Deininger, K. (1999). "Making negotiated land reform work: Initial experience from Colombia, Brazil, and South Africa." World Development 27(2):651-72.

Deininger, K. and G. Feder. (1998)."Land institutions and land markets." World Bank Policy Research Working Paper 2014.

Deininger, K. and L. Squire. (1998). “New ways of looking at old issues: Inequality and growth.” Journal of Development Economics 57():259-287.

57 Deininger, K., F. Lara, Jr., M. Maerten, and A. Quisumbing. (1999). "Agrarian reform in the Philippines: Past impact and future challenges." Processed.

De los Angeles, M. (1989). Economic Analysis of Resource Conservation by Upland

Farmers in the Philippines. Unpublished Ph. D. dissertation. Quezon City: University of the Philippines, Diliman.

De los Angeles, M. (1994). "Sustaining resource use in upland and coastal communities: Macro and micro analyses revisited." Journal of Philippine Development 38:471-518.

Demsetz, H. (1967). “Toward a theory of property rights.” American Economic Review 57: 347-359.

Department of Environment and Natural Resources (1999). "DENR position on the logging ban issue." Press release (htttp/www.denr.gov.ph).

Deininger, K. and G. Feder (1998). “Land institutions and land markets.” World Bank Policy Research Working Paper 2014.

Elazegui, D. and M. Paunlagui (1999). "Governance of fishing rights in the Philippines." Working Paper No. 98-07. Institute of Strategic Planning and Policy Studies, University of the Philippines, Los Baños.

58 Eswaran, M. and A. Kotwal (1985). “A theory of contractual structure in agriculture.”

American Economic Review 75(2):352-67.

Fafchamps, M. and B. Helms. (1996). “Local demand, investment multipliers, and industrialization: theory and application to the Guatemalan highlands.” Journal of

Development Economics, 49:61-92.

Faruqee, R and K. Carey (1997). “Land markets in South Asia: What have we learned?” World Bank Policy Research Working Paper 1754.

Feeney, D. (1988). “The development of property rights in land: A comparative study.”

Toward a Political Economy of Development: A Rational Choice Perspective. R. Bates, ed. Berkely: University of California Press.

Fields, B. (1989). “The evolution of property rights.” Kyklos 42(3):319-345.

Firmin-Sellers, K. and P. Sellers (1999). “Expected failures and unexpected successes of land titling in Africa.” World Development 27(7):1115-1128.

Floro, S. and P. Yotopoulos (1991). Informal Credit Markets and the New Institutional

Economics. Boulder: Westview Press.

59 Francisco, H. (1994). "Impact evaluation of the Central Visayas Regional Project: The upland agriculture component." Impact evaluation of the Central Visayas Regional

Project Phase I (CVRPI). L. Penalba et. al., eds. PIDS Discussion Paper Series 94-22.

Garilao, E. (1998). “The Ramos legacy in agrarian reform: A transition report.” Department of Agrarian Reform. Processed.

Geron, M. (1994). “The impact of comprehensive agrarian reform program (carp) on the crop sector.” PIDS Discussion Paper Series No. 94-15.

Gordoncillo, P. et. al. (1998). “Study on the impact of land use conversion on CARP.” Reported prepared for the Department of Agrarian Reform under the UNDP-SARDIC Program.

Grepperud, S. (1997). "Poverty, land degradation, and climatic uncertainty." Oxford

Economic Papers 49(4): 586-608.

Grossman, S. and O. Hart (1986). “The costs and benefits of ownership: a theory of vertical and lateral integration.” Journal of Political Economy 94:691-719.

Habito, C. (1989). "Fiscal policies in Philippine agriculture." Policy Considerations for

Structural Changes and Development in Philippine Agriculture. College, Laguna: University of the Philippines at Los Baños Agricultural Policy Research Program.

60

Hallagan, W. (1978). “Self-selection by contractual choice and the theory of sharecropping.” Bell Journal of Economics 9(2):344–354.

Hardin, G. (1968). "The tragedy of the commons." Science I62, 1243-48.

Hart, O. and J. Moore (1990). "Property rights and the nature of the firm." Journal of

Political Economy 98(6):1119-58.

Hayami, Y., and K. Otsuka (1993a). The Economics of Contract Choice: An

Agrarian Perspective. Oxford: Clarendon Press.

Hayami, Y., and K. Otsuka (1993b). "Kasugpong in the Philippine Rice Bowl: The emergence of new labor institutions after land reform." The Economics of Rural

Organizations: Theory, Practice, and Policy. Hoff, K. A. Braverman, and J. Stiglitz, eds. New York: Oxford University Press.

Heltberg, R. (1998). “Rural market imperfections and the farm size–productivity relationship: Evidence from Pakistan.” World Development 26(10):1807-1826.

Hyde, W., G. Amacher, and W. Magrath (1996). "Deforestation and forest land use: Theory and evidence." World Bank Research Observer 11(2):223-48.

61 Israel, D. (1997). "Overfishing in the Philippine commercial marine fisheries sector." PIDS Discussion Paper 97-07.

Israel, D. and R. Roque (1999). "Toward the sustainable development of the fisheries sector: An analysis of the Philippine Fisheries Code and the Agriculture and Fisheries Modernization Act." PIDS Discussion Paper 97-01.

Johansen, P. and K. Lofgren. (1985). The Economics of Forestry and Natural Resources. Oxford: Basil Blackwell.

Katon, B., R. Pomeroy, and A. Salamanca (1997). “The marine conservation project in San Salvador: A case study of fisheries co-management in the Philippines.” Fisheries CoManagement Working Paper No. 33. Manila: International Center for Living Aquatic Resources Management (ICLARM).

Katon, B., M. Ring, R. Pomeroy, B. and L. Gaces (1997). “Measuring co-management.” ICLARM and Haribon Foundation, Inc., Philippines. Fisheries Co-management Research Project.

Krugman, P. (1993). “Toward a counter-counterrevolution in development theory.”

Proceedings of the World Bank Annual Conference on Development Economics 1992, 15-38.

62 Kuperan, K, N. Mustapha, R. Abdullah, R. Pomeroy, E. Genio, and A. Salamanca. “Measuring transaction costs of fisheries co-management.” Manila: ICLARM.

Institute of Agrarian Studies-University of the Philippines at Los Baños. (1994). “An analysis of the effects of CARP on agricultural productivity.” Processed.

Institute of Agrarian Studies – University of the Philippines at Los Baños. (1996). “Agrarian reform beneficiaries performance monitoring and evaluation system: final report.” Processed.

Lanjouw, J. (1999). "Information and operation of markets: Tests based on a general equilibrium model of land leasing in India." Journal of Development Economics 60L497527.

Li, H. and H. Zhou (1998). "Income inequality is not harmful for growth: Theory and evidence." Review of Development Economics 2(3):318-334.

Lin, J. (1992). “Rural reforms and agricultural growth in China.” American Economic

Review 82(1):34-51.

63 Llanto, G., and C. Dingcong. “Bank behavior in an agrarian reform regime.” Financial

Intermediation in an Agrarian Reform Regime. Pasig City: Agricultural Credit Policy Council.

Llanto, G. and B. Estanislao. (1993). “The comprehensive agrarian reform program and the collateral value of agricultural lands.” Agribusiness System Assistance Program Publication No. 1.09.

McCarthy N., E. Sadoulet, and A. de Janvry (1998). "Common pool resource appropriation under costly cooperation." Processed.

Medalla, F. and M. Centeno. (1994). “Land use, urbanization, and the land conversion issue.” Spatial Development, Land Use, and Urban-Rural Linkages in the Philippines. A. Balisacan, F. Medalla, and E. Pernia, eds. Pasig City: National Economic Development Authority.

MODE, Inc. and J. Riedinger. (1998). “The impact of agrarian reform and market changes on Philippine rural households: Survey results – Key indicators.” Quezon City: MODE, Inc. Processed.

Muthoo, A.(1998). “Renegotiation-proof tenurial contracts as screening mechanisms.”

Journal of Development Economics 56(1): 1–26.

64 Newberry, D. And J. Stiglitz (1979) “Sharecropping, risk sharing, and the importance of imperfect information.” J. Roumasset et. al, eds. In Risk, Uncertainty, and Agricultural

Development. College, Laguna: Southeast Asian Regional Center for Graduate Study and Research in Agriculture (SEARCA) and Agricultural Development Council (ADC).

Ostrom, E., R. Gardner, and J. Walker (1994). Rules, Games, and Common-Pool

Resources. Boulder, Colo.:Westview Press.

Ostrom, E. (1994). “Neither market nor state: Governance of common-pool resources in the twenty-first century.” Presented at the IFPRI Lecture Series. Washington, D.C.: International Food Policy Research Institute, June 2.

Otsuka, K. and Y. Hayami (1988). “Theories of share tenancy: A critical survey.”

Economic Development and Cultural Change 37(1): 31-68.

Otsuka, K. (1991). "Determinants and consequences of land reform implementation in the Philippines." Journal of Development Economics 35(2):339-355.

Otsuka, K., V. Cordova, and C. David (1992). "Green revolution, land reform and household income distribution in the Philippines." Economic Development and Cultural

Change 40(4):719-742.

65 Otsuka, K. (1998). “ Land tenure and management of land and trees in Asia and Africa.” Paper prepared for the Policy Seminar at the International Food Policy Research Institute, 3 September.

Otsuka, K. (1999). “Land tenure, land reform, and rural poverty in Asia.” Processed.

Park, A. and B. Johnston (1995). “Rural development and dynamic externalities in Taiwan’s Structural Transformation.” Economic Development and Cultural Change 44(1):181-208.

Peñalba, L., E. Escueta, and B. Pantoja. (1996). “Benchmark Survey of the Agrarian Reform Communities: Final Report.” Institute of Agrarian Studies – University of the Philippines at Los Baños. Processed.

Platteau, J. (1996). “The evolutionary theory of land rights as applied to sub-Saharan Africa: A critical assessment.” Development and Change 27(1):29-26.

Pomeroy, R., B. Katon, and I. Harkes. (1997). “Fisheries co-management: Key conditions and principles drawn from Asian experiences.” International Center for Living Aquatic Resources Management (ICLARM).

Ranis, G., and F. Stewart (1993). “Rural nonagricultural activities in development: theory and applications.” Journal of Development Economics, 40(1):75-101.

66

Ray, T. (1999). “Share tenancy as strategic delegation.” Journal of Development

Economics 58(1):45-60.

Seabright, P. (1993). "Managing local commons: Theoretical issues in incentive design."

Journal of Economic Perspectives 7(4):113-134.

Sethi, R. and E. Samanathan (1996). "Evolution of social norms in common property use." American Economic Review 86(4):766-788.

Skinner, J. (1993). "If agricultural land taxation is so efficient, why is it so rarely used?"

The Economics of Rural Organizations: Theory, Practice, and Policy. Hoff, K. A. Braverman, and J. Stiglitz, eds. New York: Oxford University Press.

Stiglitz, J. (1974). “Incentives and risk sharing in sharecropping.” Review of Economic

Studies 41(2):219-256.

Stiglitz, J. and A. Weiss. ”Credit rationing in markets with imperfect information.”

American Economic Review 71(3):393-410.

World Bank (1999). Philippines Tree Crops for Rural Development: Issues and

Strategies. vol. 2. World Bank: Rural Development and Natural Resource Sector Unit Report No. 19281-PH.

67

Appendix

The Survey Frames of the Major Data Sets The Benchmark Survey, covering crop year 1989-1990, was aimed at providing a basis for comparing the progress of agrarian reform at the farm and household level. Over 8,000 households were drawn from 400 barangays out of 41 provinces having the highest proportions of lands subject to CARP. Sample selection was also stratified by ecological zone (i.e., upland, lowland and coastal zones). The Benchmark Survey of ARCs, like the Benchmark Survey, included both ARBs and non-ARBs among the respondents. Data pertained to crop year 1993-1994. This survey covered 61 ARCs with 3,656 respondents (approximately 60 each ARB). Within each ARC, barangays were stratified by ecological zone when possible. Selection of ARBs and non-beneficiaries was done by proportional sampling. The MODE Impact of Agrarian Reform survey (henceforth the MODE survey) consisted of interviews of a subset of respondents (around 1,500 in all) from the Benchmark survey. Its sampling design hewed closely to that of the Benchmark Survey. Unfortunately, the survey instrument diverged greatly from that used in the Benchmark survey, making the usefulness of the MODE survey for longitudinal comparison limited. The sample design of PMES was also based on the Benchmark Survey, but its coverage was limited to ARBs under DAR jurisdiction. Phase 2 was a pilot test of the PMES covering 3,411 ARB respondents from 20 provinces. Selection of provinces applied island group stratification (Luzon, Visayas, and Mindanao) and size stratification (large and small provinces). Meanwhile Phase 3 is yet to be finalized, although summary findings are available. This survey is distinguished by its validity for regional analysis as well as its inclusion of ARCs as a distinct domain. The ARBs are first stratified in terms of residence or nonresidence in an ARC. The selection of ARC barangays from each included province is based on categories of LAD accomplishment. For non-ARC residents, selection from

69

each included province is based on a subset of municipalities, followed by a subset of barangays, and finally a subset of ARBs in the barangay. Finally, the longitudinal study of Deininger et al. (1999) utilized villages that had been surveyed by the International Rice Research Institute as well as the International Food Policy Research Institute in 1985, 1989, and 1998. The 1989 survey was able to collect information on inheritances, assets and the history of land transactions. Five villages were covered: two in Central Luzon and three in Panay island. In each area, the sample included one village with irrigated rice land and a favorable agro climatic environment, as well as one village with rainfed production combined with supplemental irrigation. The

fifth

village

(in

Panay)

had

an

unfavorable

uplandenvironment.